New Home!

This project has been moved to: https://github.com/dialogflowai/dialogflow

简体中文 | English

Dialog flow chat bot

only ONE executable file, you can use it directly, including intent detection, AI management, a visual process editor and a response system.

✨ Features

• 🛒 Light Only ONE executable file, it can run smoothly on laptops without GPUs (data files will be created at runtime automatically).
• 🐱‍🏍 AI powered Integrated Huggingface local models (Llama, Phi-3, Gemma, Multilingual E5, MiniLM L6v2, NomicEmbedTextV1_5, etc.), Ollama and OpenAI, this can be used for Chat, Text generation and Intent detection.
• 🚀 Fast Built on Rust and Vue3.
• 😀 Simple Use the mouse to drag and drop with our intuitive node-based editor.
• 🔐 Safe 100% open source, all runtime data is saved locally (Using OpenAI API may expose some data).

Give it a try!

• 🐋 Docker We provided an image on Docker Hub at dialogflowchatbot/demo
• 💻 Binary releases, please check here

By default application will listen to 127.0.0.1:12715, you can use -ip and -port specify new value, e.g.: dialogflow -ip 0.0.0.0 -port 8888

Check out introduction page

https://easyflow-ai.github.io/

Function nodes

Using the different nodes above, to arrange and combine, you can get a conversational bot that can handle problems in different scenarios.

Screenshots

Trying a demo dialog flow

Setup a condition branch

Text generation

Testing a dialog flow

Get started

Docker image

1. docker pull dialogflowchatbot/demo
2. docker run -dp 127.0.0.1:12715:12715 --name dialogflowdemo dialogflowchatbot/demo
3. Open your browser and visit: http://127.0.0.1:12715/

Binary release

1. From Github release page, depending on the operating system, download the application.
2. Run it directly, or use the -ip and -port parameters to perform the listening IP address and port, e.g.: dialogflow -ip 0.0.0.0 -port 8888.
3. Open your browser and visit http://localhost:12715 (by default) or http://new IP:new port to see the application in action
4. Add a main flow and click its name into it
5. Create dialog flow by dragging and drop nodes onto canvas
6. Test it